Analytical Biochemistry 451 (2014) 56–62
Contents lists available at ScienceDirect
Analytical Biochemistry journal homepage: www.elsevier.com/locate/yabio
Gold nanostructures for the multiplex detection of glucose-6-phosphate dehydrogenase gene mutations Nianjia Seow a, Poh San Lai b,⇑, Lin-Yue Lanry Yung a,⇑ a b
Department of Chemical and Biomolecular Engineering, Faculty of Engineering, National University of Singapore, Singapore 119260, Singapore Department of Pediatrics, National University Health System, Yong Loo Lin School of Medicine, National University of Singapore, Singapore 119228, Singapore
a r t i c l e
i n f o
Article history: Received 19 October 2013 Received in revised form 18 January 2014 Accepted 21 January 2014 Available online 1 February 2014 Keywords: Gold nanoparticles Glucose-6-phosphate dehydrogenase (G6PD) deficiency Mutation screening Multiplex detection Gel electrophoresis
a b s t r a c t We describe a gold nanoparticle-based technique for the detection of single-base mutations in the glucose-6-phosphate dehydrogenase (G6PD) gene, a condition that can lead to neonatal jaundice and hemolytic anemia. The aim of this technique is to clearly distinguish different mutations frequently described within the Asian population from their wild-type counterparts and across different mutant variants. Gold nanoparticles of different sizes were synthesized, and each was conjugated with a single-strand DNA (ssDNA) sequence specific for a particular mutation in the G6PD gene. It was found that only mutant targets presented a characteristic band on the agarose gel, indicating the successful formation of dimeric nanostructures. No such dimer bands were observed for the wild-type targets. The difference in the relative dimer band levels allowed different mutant variants to be distinguished from one another. The technique was further validated using G6PD-deficient patient samples. This simple mutation detection method with direct result readout is amenable for rapid and mass screening of samples. Ó 2014 Elsevier Inc. All rights reserved.
Many genetic diseases are caused by single nucleotide point mutations, for example, cystic fibrosis and thalassemia [1]. In particular, glucose-6-phosphate dehydrogenase (G6PD)1 deficiency is a recessive genetic condition arising from single-base mutations in the G6PD gene on the X chromosome and affects 200 to 400 million individuals worldwide [2–4]. It is estimated that 7.5% of global populations carry one or two G6PD mutant alleles, with 2.9% being deficient for this enzyme deficiency. It is also thought that this condition is highly prevalent in malaria endemic countries because it confers resistance against malaria [2]. G6PD-deficient subjects are usually asymptomatic until triggered by infections or the ingestion of certain foods or drugs such as anti-malarial drugs. Clinical outcomes commonly described include neonatal jaundice and acute hemolytic anemia [5]. Thus, there is a need to detect G6PD-deficient subjects in order to avoid triggers for hemolytic anemia and use a substitute medication course when necessary [6]. Routine screening methods typically involve colorimetric biochemical assays. However, ⇑ Corresponding authors. E-mail addresses: [email protected], [email protected] (P.S. Lai), [email protected] (L.-Y.L. Yung). 1 Abbreviations used: G6PD, glucose-6-phosphate dehydrogenase; nAu, gold nanoparticle; PCR, polymerase chain reaction; ssDNA, single-strand DNA; TEM, transmission electron microscopy; HAuCl4, hydrogen tetrachloroaurate(III) trihydrate; Na3Ct, trisodium citrate dehydrate; PPBS, 4,40 -(phenylphosphinidene)bis (benzenesulfonic acid) dipotassium salt hydrate 97%; TBE, Tris–borate–EDTA; UV, ultraviolet; DLS, dynamic light scattering; OEG, oligo(ethylene glycol). http://dx.doi.org/10.1016/j.ab.2014.01.014 0003-2697/Ó 2014 Elsevier Inc. All rights reserved.
although these methods are rapid and relatively cheap, they are qualitative or semi-quantitative and cannot identify the exact mutations [7]. They are also not sensitive enough to detect heterozygotes where both normal and G6PD-deficient erythrocytes are present [8,9]. Molecular diagnosis is possible through identification of the mutations in the G6PD gene. Such mutations have been reported at various locations of the gene in affected individuals, and the dominance of particular mutations is geographically linked. Specific mutation genotypes have been associated with different populations. For example, c.95A>G, c.871G>A, c.1004C>T, c.1024C>T, c.1376G>T, and c.1388G>A mutations are especially common among the Chinese population; these mutations are commonly referred to as the Gaohe, Viangchang, Fushan, Chinese-5, Canton, and Kaiping variants, respectively [10–13]. Typically, molecular diagnosis for any genetic disorder involves either a mutation screening method followed by confirmation through DNA sequencing or directly proceeding to sequencing. This is often laborious due to the experimental steps involved and the sequence data analysis required. A rapid approach for diagnosis is through genotyping for mutations common to a geographical locale, followed by subsequent testing for more uncommon and rare mutations if the former is not informative. Thus, given the clinical relevance and the complement of mutations associated with this condition, there is a motivation to develop a simple and inexpensive multiplex assay with direct result readout for large-scale or population-based screening of G6PD, especially in developing countries where G6PD deficiency is prevalent.
Gold nanostructures for mutation detection / N. Seow et al. / Anal. Biochem. 451 (2014) 56–62
The advent of nanotechnology has led to improvements in molecular techniques in areas such as sensitivity, selectivity, and immediacy of readouts. Nanoparticles offer unique properties, such as surface plasmon resonance, and are amenable to biofunctionalizations, making them ideal as labeling tags or readout platforms [14–16]. Mirkin and coworkers have developed various colorimetric assays that leverage on the spectral shift of gold nanoparticles (nAu) as they aggregate in the presence of DNA targets [17]. The readouts can easily be visualized by the naked eye. Further application using bio-barcode assay even offers polymerase chain reaction (PCR)-like performance with a zeptomolar (zM) detection limit [17,18]. Control in the functionalization of nAu was also described by Alivisatos and coworkers as they showed the formation of discrete nAu structures bearing a distinct number of single-strand DNA (ssDNA) that is evidenced in transmission electron microscopy (TEM) and gel platforms [19,20]. This allows the formation of distinct nanostructures, as Qin and Yung demonstrated through the use of nAu dimers to discriminate single-nucleotide mutations in diseases such as Duchenne muscular dystrophy [21–23]. Here we report the use of different nAu probes for detecting four characteristic G6PD mutations prevalent in the Asian population, namely Canton, Mahidol, Union, and A+ variants [3,24,25]. These probes are not only specific for a perfectly matched mutant target relative to the wild-type nontarget, but when used in tandem they are also able to discriminate variants in the G6PD gene. This multiplex readout is presented in a direct and unambiguous manner, which gives a yes/no readout for the variant nucleotide of interest and also discriminates among different single-base mutations through the differential mobility of the probes in agarose gel electrophoresis. Clinical samples were probed with different types and sizes of probes, and results had conformed to predicted expectations. The success of this technique has wider implications in the screening for disease-associated single-nucleotide mutations and polymorphisms and for biomarker discovery in general. Materials and methods Materials Hydrogen tetrachloroaurate(III) trihydrate (HAuCl4), trisodium citrate dihydrate (Na3Ct), tannic acid, 4,40 -(phenylphosphinidene)bis (benzenesulfonic acid) dipotassium salt hydrate 97% (PPBS), NaCl, MgCl2, and Tris–borate–EDTA (TBE) buffer were purchased from Sigma–Aldrich (Singapore). Tris buffer (pH 8.0) and agarose powder were purchased from 1st Base (Singapore). Modified synthetic DNA molecules (modified with or without 30 or 50 thiol end group) were purchased from 1st Base Custom Oligos (Singapore). Milli-Q water with resistance > 18 MO/cm was used throughout the experiments. For the amplification of patient samples, standard PCR conditions were applied using a thermocycler (Biometra, Germany): initial denaturation at 95 °C for 7 min; followed by 35 cycles of 95 °C for 1 min, 58 °C for 1 min, and 72 °C for 3 min; and a final extension at 72 °C for 7 min. Amplifications were performed in total reaction volumes of 25 ll containing 100 ng of DNA, 0.4 lM forward and reverse primers, 1 U of Taq DNA Polymerase, 0.4 mM deoxynucleotide triphosphates (dNTPs), 2 mM MgCl2, and 1 buffer (Promega, Singapore). All samples were obtained following approved institutional ethics guidelines, and the study was approved by the institutional review board (IRB) of the National University of Singapore.
57
precursor, synthesis time, and temperature) were kept constant, whereas concentration of reductant (either Na3Ct only or both Na3Ct and tannic acid) was varied [27,28]. nAu products were characterized first with ultraviolet (UV)–visible spectroscopy and TEM (JEM-2100F, 200 kV, Joel). The TEM micrographs were analyzed with ImageJ software (National Institutes of Health) for size measurement (P200 particles). The particle size distribution was further characterized by determining particles’ electrophoretic mobilities on a 3% agarose gel (5 V/cm) and dynamic light scattering (DLS) (Zetasizer Nano ZS, Malvern). Fabrication of nAu–ssDNA conjugate probes Surface passivation of nAu with PPBS was first done for at least 4 h, followed by incubation with ssDNA specific for the mutant targets. Here, 100- and 18-base ssDNA–nAu conjugate probes for each target were prepared (the gene sequences of all the ssDNA can be found in the online Supplementary material). Next, 3 h after the oligo addition, thiolated oligo(ethylene glycol) (OEG) was added at a 500:1 OEG/nAu ratio. The nAu–ssDNA probes were prepared in 75 to 200 mM NaCl initially (depending on the size of nAu), and gradually the NaCl concentration was raised to between 350 and 600 mM (larger nAu conjugates are generally less stable at higher NaCl conditions). The probes were subsequently washed in 50 mM Tris and NaCl. The 100-base probes were run on a 3% agarose gel, and conjugate monomers (probes bearing a single ssDNA strand) were recovered. The 18-base probes were used without prior purification. The concentrations of probes were estimated from their UV absorbance [29,30]. Prior to use, all probes were screened using UV spectroscopy and DLS to ensure that they remained monodispersed and unaggregated. Probes without targets were also used as controls in subsequent tests and gel runs, and their monodispersity was further confirmed by the (single) band observed. Formation of dimeric nanostructures in the presence of mutant targets Based on our previous work, and after optimization for different sizes of nAu, a probe/target ratio ranging from 1:1 to 1:1.5 was used. By default, 1 pmol of synthetic DNA was used as a target for all of the runs. Hybridization conditions were kept at 50 mM Tris buffer (pH 8.0), 100 mM NaCl, and 2 mM MgCl2, with the samples being heated up to 75 °C and cooled by 0.5 °C every 30 s. The hybridized samples were left to stand for at least 1 h. The samples were then characterized using a 3% agarose gel (75 V for 45 min at 4 °C). Gel images were taken using the Syngene GeneGenius UV/ white light gel documentation system. The gel images, with a 10% enhancement of contrast, were subsequently analyzed using ImageJ software, and the density of the dimer bands was measured relative to that of the unbound probes control. This densitometric analysis provided more quantitative information on the dimer bands and further confirmed the presence of the target. A schematic of the nanostructure formation and gel readout is shown in Fig. 1, where two nAu–ssDNA probes (carrying 100and 18-base ssDNA, respectively) hybridized to a mutant target sequence and formed a dimeric gold nanostructure. The successful formation of the structure was visualized by agarose gel electrophoresis, where a distinct band could be readily seen by the naked eye. Any sequence mismatches would not lead to the formation of such dimer bands. Results
Synthesis and characterization of gold nanoparticles
Characterization of different sizes of gold nanoparticles
The synthesis of nAu was based on the Turkevich protocol [26] for the synthesis of colloidal gold. Reaction conditions (of HAuCl4
The combination of water, Na3Ct, and HAuCl4 is integral to the successful synthesis of nAu of the desired size(s) because varying
58
Gold nanostructures for mutation detection / N. Seow et al. / Anal. Biochem. 451 (2014) 56–62
wells
Dimers
& Unbound probes
Probes only
Probes with perfectmatched target
Fig.1. Design of nAu–ssDNA probe system. A 100- and 18-base probe tandem would bind onto the same single-stranded target to form a dimer. The mutation site is located on the 18-base probe. The 3% agarose gel image shows the formation of dimers in the presence of a perfect matched target. In the absence of targets (probes only), no such dimer bands were observed.
the Na3Ct/HAuCl4 ratio changes the reaction pathway and, in turn, affects the quality of the nAu product. Peng and coworkers reported that, besides acting as a reducing and capping agent, citrate is a pH mediator, and controlling its addition into the reaction mixture can have a direct effect on the synthesis outcome [31]. In addition, fixing the Na3Ct/HAuCl4 ratio at approximately 3 produced the smallest nanoparticle size. When the Na3Ct/HAuCl4 ratio decreased from 3 or below, the product size was progressively larger because the conditions promoted the growth of nanocrystals in the traditional route of nucleation followed by diffusion-controlled net growth [32]. When the ratio increased above 3, there was a corresponding increase in product size, although the increase was more gradual. This could be attributed to the high citrate concentration, which caused the reaction pH to be lower than the pH 6.5 switch point. As a result, there were changes to the structure and reactivity of the Au(III) complexes, with the reaction pathway instead involving nucleation, aggregation, and intraparticle ripening, which led to larger sized nAu. Given that we were leveraging on the properties of different nAu sizes, it was imperative that we could properly control the sizes of nAu synthesized and replicate them with ease. Hence, we fixed the starting concentration of HAuCl4 at 0.1 mg/ml and varied the concentration of Na3Ct at a fixed total reaction volume and synthesis conditions such as temperature and incubation time. Na3Ct was at least five times in excess of HAuCl4. Through controlling the relative concentrations of HAuCl4 precursor and reductant, the resulting nAu showed a size range of 15 to 30 nm. To achieve nAu sizes of 12 nm or lower, a combination of Na3Ct and tannic acid was used, with 9-nm nAu being the smallest synthesized. Size measurement via TEM and DLS further verified the results (Fig. 2A and B). The synthesized nAu showed a narrow size distribution, with the standard deviation within 10% of the average particle size. The DLS readout also showed that each synthesized size presented a peak position distinct from that of the other sizes. Under agarose gel electrophoresis, the as-synthesized nAu of different sizes exhibited the mobility corresponding to particle size, with larger particles showing slower mobilities than their smaller sized counterparts (Fig. 2A). The mobility differentiation among all particle sizes subsequently allowed us to use electrophoresis as the readout platform. The size range (9–25 nm) was also adequate for a multiplex detection scheme because the mobilities of different size bands were visually distinguishable. Formation of discrete nanostructures with different sizes of gold nanoparticles We first used the Canton G6PD point mutation, which showed a G > T mutation, as a proof of principle for the methodology. The aim was to investigate the ability of the nAu probes to form dimers
as well as the relative mobilities of the nanostructures. Different sizes of nAu (8.5, 11, 14, 20, and 25 nm) were conjugated to ssDNA specific to the mutant Canton variant. Probes bearing a single DNA strand were recovered from the agarose gel to ensure dimer formation and no other nanostructures. After optimization of hybridization conditions for each of the sizes, dimers were successfully formed and presented as distinct dimeric bands on the gel platform (Fig. 3). The 8.5-nm 18-base and 25-nm 100-base probes were used as the upper and lower limit controls because the former would show the greatest mobility and the latter would show the least mobility among the probes. Thus, they served as reference points for evaluating the mobilities of the unbound probes and dimer bands when targets were present. All of the unbound probe bands were localized within these two reference bands, with the dimer bands showing a slower mobility behind the unbound probe bands. Potentially, the formation of dimers not only would provide a yes/no readout with regard to the presence of a DNA target but also would discriminate between different mutations due to the respective dimers being presented at distinct locations on the gel. Higher order bands that migrate at slower mobility than the dimer bands were formed due to multiple ssDNA strands conjugated on the 18-base nAu probes. Because 18-base ssDNA is too short to induce sufficient electrophoretic mobility, nAu probes with different numbers of ssDNA strands could not be resolved into a distinct gel band during nAu–ssDNA probe fabrication; thus, no purification was done prior to use. Instead, the number of 18-base ssDNA on the 18-base probes was adjusted only through careful stoichiometric control, such that dimers would still be the dominant population in the presence of complementary target DNA, but the presence of dimer or higher order -mer bands was still indicative of the presence of target DNA. Multiplex detection with discrimination between mutant and wildtype targets and across different point mutations nAu–ssDNA probes for the four different variants (Union, Mahidol, Canton, and A+) were fabricated, and the dimeric nanostructures reflecting positive readouts were found to exhibit distinct electrophoretic mobilities. As seen in Fig. 4, the four variants were queried over eight nAu sizes, and dimer bands were observed only for the mutant targets to which the probes were perfectly complementary. The wild-type sequences that contained a single base difference did not produce any dimer bands. The mismatch site, being located on the shorter 18-base nAu probe, was affected by the electrostatic–steric effect exerted by the nearby nAu, and this destabilized the complementary binding of the probe for the target, which had an overlap of only 13 bases. Despite different sizes of nAu exhibiting varied surface charge densities, the electrostatic–steric effect was still significant in bringing about the desired discrimination property. A single-base mismatch was enough to destabilize and dissemble any nanostructure formed. More importantly, all of the dimer bands were observed at different positions on the gel that were distinct from one another. Although the probes carried ssDNA of the same length, the difference in the nAu sizes for different mutant targets resulted in the differential mobilities observed. This was attributed to the mass being an important property that controlled the particle mobility on a gel platform, with the larger nAu probes exhibiting slower dimer movement [33–35]. This provided an added level of discrimination for different sequences because each was expected to have a unique position based on the sizes of the nAu particles. In this particular case, the spatial multiplex detection of different G6PD variants discriminated a mutant sequence not only from its wild-type counterpart but also across different point mutations through a simple and direct observation of the mobilities of the nAu probe across lanes. A positive readout would be expected in relation to
Gold nanostructures for mutation detection / N. Seow et al. / Anal. Biochem. 451 (2014) 56–62
59
Fig.2. (A) TEM micrographs and gel image of six different sizes of nAu—9.0 nm (a), 12.3 nm (b), 15.2 nm (c), 17.3 nm (d), 21.0 nm (e), and 27.2 nm (f)—synthesized with different amounts of citrate and tannic acid reductants. (B) Particle distribution of different sizes of nAu using DLS. All of the particles show distinct size distributions and electrophoretic mobilities.
8.5nm dimer
11nm dimer
14nm dimer
20nm dimer 25nm dimer
25nm 100-base probe (lower limit control) 8.5nm 18-base probe (upper limit control)
denotes the dimermonomer separation
Direction of migration
Fig.3. Dimer formation with 8.5-, 11-, 14-, 20-, and 25-nm nAu (with Canton variant as the model) in 3% agarose gel (75 V, 60 min). Target loading was 1 pmol, and the amounts of Canton mutation-specific probes used were varied to give optimal observation of the dimer band. Higher order multimer bands were attributed to the unpurified 18-base probes, some of which may carry more than one ssDNA per nAu and led to the formation of not just dimer but also trimer, tetramer, and so forth.
60
Gold nanostructures for mutation detection / N. Seow et al. / Anal. Biochem. 451 (2014) 56–62
Fig.4. Multiplex detection of four variants (Union, Mahidol, Canton, and A+) complement on a 3% agarose gel. For each mutation type, the same quantity of the probes specific to the particular mutation was used to detect mutant and wild-type targets that showed a single-base difference. The appearance of dimer bands on the gel (or the lack thereof) provides identification to the type of target present. The dimer band for each variant shows distinct mobility from the others due to the different sizes of nAu used, providing an advantage in discriminating target detection during multiplex screening of varied G6PD mutations. mut: mutant (perfectly matched) target, expected to produce dimer bands; wt: wild-type (mismatched) target, not expected to produce dimer bands.
the presence of the specific dimer bands as well as its relative position to its counterparts. Querying of clinical samples using different sized probes To validate the above approach, clinical samples from G6PDdeficient patients were analyzed using the different types of nAu probes. Four different patient samples—two with the Mahidol mutation (M1 and M2) and two with Canton mutation (C3 and C4)—were obtained with informed consent and IRB approval. The genomic DNA samples were PCR amplified, purified, and subsequently detected by the respective Mahidol and Canton probes. From Fig. 5, when queried with their specific probes, the Mahidol (with 11-nm probes) and Canton (with 15-, 17-, and 20-nm probes) patient samples all exhibited dimer bands. When the target was absent, no dimer band was observed. The observations were also semi-quantified via the density of the bands, which took into account the intensity and area of the bands, and the dimer bands’ densities were approximately 10 to 20% that of the unbound probes. The densitometric analysis provided a semi-quantitative assay for confirming the presence of the dimers (and target). The dimer bands, although less dense than their synthetic target counterparts, were easily detected by the naked eye without the need for additional instruments. Fainter bands were observed because the clinical PCR-amplified samples were long duplex DNA, unlike the single-stranded synthetic DNA targets used in the previous sections. Given that the duplexes are typically more than 120 bases in
11nm Mahidol probes Synthetic Mahidol Patient M1 Patient M2 (Mahidol (Mahidol target mutation) mutation) control
15nm Canton probes Patient C3 (Canton mutation)
Patient C4 (Canton mutation)
length, whereas the 18-base probes have only 13 complementary bases to the target, the tendency for the target duplex to rehybridize with its complementary strand is higher. Thus, it was necessary to include an annealing step, such that duplexes would dehybridize and the target strands would be available for binding with the nAu probes, and all four clinical samples were successfully detected. We have shown that different sizes of probes, ranging from 11 to 20 nm, were equally capable of detecting the Canton and Mahidol clinical samples, hence validating the multiplex detection proposed. Potentially, this can be a useful technique for easy and direct screening, especially at the clinical setting for precise detection of common G6PD mutations. Discussion The development of nAu-based diagnostic techniques hinges much on the understanding of the nature and properties of nAu and their surface chemistries and, finally, fabricating the nAubased probes for the desired applications. For example, the desired surface loading of DNA onto the nanoparticle differs depending on the type of application, whereas care is taken to ensure the stability of nanoparticle dispersion and kinetics of DNA binding onto the nAu surface. Typically, high DNA-loaded conjugates are used for amplification purposes and aggregation assays [36]. On the other hand, ssDNA–nAu conjugates with low ssDNA loading have shown much promise as building blocks in complex nanostructure formation. However, the need for a controlled loading of ssDNA poses a
17nm Canton probes Patient C3
Patient C4
20nm Canton probes Patient C3
Patient C4
Control (no target)
Dimers (probes bound to target)
Unbound probes Dimers expected (yes/no) Dimers formed (yes/no) Relative intensity to unbound probes
yes
yes
yes
yes
yes
yes
yes
yes
no
yes
yes
yes
yes
yes
yes
yes
yes
no
0.17
0.13
0.18
0.09
0.09
0.21
0.12
0.15
0
Direction of migration
Fig.5. PCR-amplified samples from four patients—two with Mahidol mutation (M1 and M2) and two with Canton mutation (C3 and C4)—were received. The patient samples were queried with the respective mutation-specific probes (11-nm Mahidol probes for M1 and M2; 15-, 17-, and 20-nm Canton probes for C3 and C4). All of the patient samples exhibited dimer bands when tested with their respective probes, which followed predicted expectations. The readout is further characterized via the density of the dimer bands relative to the unbound probes. In the absence of targets, no dimers were observed.
Gold nanostructures for mutation detection / N. Seow et al. / Anal. Biochem. 451 (2014) 56–62
challenge to probe fabrication, which affects the formation of distinct nanostructures and their use in plasmonic and electronic applications. Gel electrophoresis and anion-exchange high-performance liquid chromatography (HPLC) have been the foremost choices to separate conjugates bearing different numbers of ssDNA due to their compatible resolution and ease of use. In particular, the gel platform was found to be suitable not just for the purification and subsequent recovery of conjugates bearing distinct numbers of ssDNA but also for the visualization of different nanostructures formed achieved in this work. Much care was taken during the probe fabrication process because inherently nAu are sensitive to environmental conditions such as Na+ levels, which might bring forth aggregation. Thus, for the timely fabrication of the desired conjugate probes, addition of NaCl was carefully done because it is paradoxically known that a higher NaCl level can enhance DNA binding onto the nAu surface while at the same time promoting clustering of nAu and eventual aggregation, handicapping the nanostructure formation [37]. In our case, the ssDNA conjugation protocol was optimized for all of the nAu sizes used, especially the larger sized nAu particles that were more susceptible to aggregation at high salt loadings relative to their smaller sized counterparts. This ensured the efficient hybridization of probes to the targets without interference of salt-induced clustering of probes or formation of nonspecific nanostructures. Contrary to our previous work, where thiolated 5-thymidine (5T) was used in the preparation of the conjugate probes, this study showed that OEG was as good a passivating agent as 5T while being only a tenth in cost. OEG was able to compete with thiolated ssDNA and allowed both the 100- and 18-base ssDNA coverage on the nAu to be kept low, thereby favoring the formation of discrete conjugates with low ssDNA loading [38]. From the hybridization studies, probe–target binding was just as effective, which suggested that the OEG-coated surface was non-interacting and kept the probe ssDNA extended outward and not collapsed onto the nAu surface, hence facilitating the hybridization process. Although we tested eight nAu sizes in the 7- to 25-nm range, it is possible to include additional nAu sizes as small as 5 nm and as large as 40 nm, with each size being used to detect for a specific mutant variant of the G6PD gene. This will be especially useful to simultaneously cover screening of other less common point mutations in the G6PD gene. This, coupled with the developed procedure of probe fabrication, target hybridization, and visualization as described in this study, offers a potentially powerful multiplex detection method for G6PD screening of known mutations in a population. As compared with conventional molecular techniques and other nanoparticle-based platforms, gold nanostructure probes provide a simple and direct manner of detection that can be applied to the detection of point mutations in G6PD deficiency as well as for other disorders.
Conclusion Given the need for a simple and direct system for the detection of multiple nucleic acid targets, we have developed an nAu system that, through the formation of distinct structures from different sizes of nAu, can be clearly visualized on a gel platform. Wild-type and single-base mutant sequences were discriminated, as well as across variants, and this technique was further validated using clinical samples. The ability to detect the readouts from the nAu hybridization by direct visualization from a gel-based assay makes this a simple and inexpensive method. Our method can potentially be applied for screening other pathogenic point mutations in disease genes as well as for genotyping common single-nucleotide polymorphisms used as markers for disease association studies in disease gene discovery.
61
Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.ab.2014.01.014.
References [1] M.N. Kuppuswamy, J.W. Hoffmann, C.K. Kasper, S.G. Spitze, S.L. Groce, S.P. Bajaj, Single nucleotide primer extension to detect genetic diseases: experimental application to hemophilia B (factor IX) and cystic fibrosis genes, Proc. Natl. Acad. Sci. USA 88 (1991) 1143–1147. [2] M.D. Cappellini, G. Fiorelli, Glucose-6-phosphate dehydrogenase deficiency, Lancet 371 (2008) 64–74. [3] P. Delbini, L. Duca, I. Nava, D. Tavazzi, M.D. Cappellini, Gene symbol: G6PD. Disease: glucose-6-phosphate dehydrogenase deficiency, Hum. Genet. 124 (2008) 312–313. [4] P.S. Lai, T. Shirakawa, K. Nishiyama, M. Matsuo, R. Joseph, S.H. Quak, G6PD deficiency and application of the MPTP technique, in: P.S. Lai, E.P.H. Yap (Eds.), Frontiers of Human Genetics: Diseases and Technologies, World Scientific, Singapore, 2001, pp. 121–138. [5] J.E. Frank, Diagnosis and management of G6PD deficiency, Am. Fam. Physician 72 (2005) 1277–1282. [6] W.E. Evans, M.V. Relling, Pharmacogenomics: translating functional genomics into rational therapeutics, Science 286 (1999) 487–491. [7] I.S. Tantular, F. Kawamoto, An improved, simple screening method for detection of glucose-6-phosphate dehydrogenase deficiency, Trop. Med. Int. Health 8 (2003) 569–574. [8] G.J. Reclos, C.J. Hatzidakis, K.H. Schulpis, Glucose-6-phosphate dehydrogenase deficiency neonatal screening: preliminary evidence that a high percentage of partially deficient female neonates are missed during routine screening, J. Med. Screen. 7 (2000) 46–51. [9] O. Ainoon, A. Alawiyah, Y.H. Yu, Semiquantitative screening test for G6PD deficiency detects severe deficiency but misses a substantial proportion of partially-deficient females, Southeast Asian, J. Trop. Med. Public Health 34 (2003) 405–414. [10] L. Li, Y.Q. Zou, Q.Z. Xiao, T.Z. Yan, X.M. Xu, Development and evaluation of a reverse dot blot assay for the simultaneous detection of six common Chinese G6PD mutations and one polymorphism, Blood Cells Mol. Dis. 41 (2008) 17– 21. [11] A. Minucci, L. Gentile, C. Zuppi, B. Giardina, E. Capoluongo, Rapid and simple identification of the commonest glucose-6-phosphate dehydrogenase (G6PD) Italian mutations: from DNA extraction to genotyping, Clin. Chim. Acta 413 (2012) 1018–1019. [12] A. Minucci, P. Concolino, M. Antenucci, C. Santonocito, F. Ameglio, C. Zuppi, B. Giardina, E. Capoluongo, Description of a novel missense mutation of glucose6-phosphate dehydrogenase gene associated with asymptomatic high enzyme deficiency, Clin. Biochem. 40 (2007) 856–858. [13] T. Yan, R. Cai, O. Mo, D. Zhu, H. Ouyang, F. Huang, L. Li, X. Liang, X. Xu, Incidence and complete molecular characterization of glucose-6-phosphate dehydrogenase deficiency in the Guangxi Zhuang autonomous region of southern China: description of four novel mutations, Haematologica 91 (2006) 1321–1328. [14] S.S. Agasti, S. Rana, M.H. Park, C.K. Kim, C.C. You, V.M. Rotello, Nanoparticles for detection and diagnosis, Adv. Drug Deliv. Rev. 62 (2010) 316–328. [15] J.N. Anker, W.P. Hall, O. Lyandres, N.C. Shah, J. Zhao, R.P. Van Duyne, Biosensing with plasmonic nanosensors, Nat. Mater. 7 (2008) 442–453. [16] Z. Zhan, X. Ma, C. Cao, S.J. Sim, Gold-based optical biosensor for singlemismatched DNA detection using salt-induced hybridization, Biosens. Bioelectron. 32 (2012) 127–132. [17] M.S. Han, A.K.R. Lytton-Jean, B.K. Oh, J. Heo, C.A. Mirkin, Colorimetric screening of DNA-binding molecules with gold nanoparticle probes, Angew. Chem. Int. Ed. Engl. 45 (2006) 1807–1810. [18] J.M. Nam, S.I. Stoeva, C.A. Mirkin, Bio-bar-code-based DNA detection with PCRlike sensitivity, J. Am. Chem. Soc. 126 (2004) 5932–5933. [19] D. Zanchet, C.M. Micheel, W.J. Parak, D. Gerion, S.C. Williams, A.P. Alivisatos, Electrophoretic and structural studies of DNA-directed Au nanoparticle groupings, J. Phys. Chem. B 106 (2002) 11758–11763. [20] D. Zanchet, C.M. Micheel, W.J. Parak, D. Gerion, A.J. Alivisatos, Electrophoretic isolation of discrete Au nanocrystal/DNA conjugates, Nano Lett. 1 (2001) 32– 35. [21] W.J. Qin, L.Y.L. Yung, Nanoparticle–DNA conjugates bearing a specific number of short DNA strands by enzymatic manipulation of nanoparticle-bound DNA, Langmuir 21 (2005) 11330–11334. [22] W.J. Qin, L.Y.L. Yung, Nanoparticle-based detection and quantification of DNA with single nucleotide polymorphism (SNP) discrimination selectivity, Nucleic Acids Res. 35 (2007) e111. [23] W.J. Qin, O.S. Yim, P.S. Lai, L.Y.L. Yung, Dimeric gold assembly for detection and discrimination of single nucleotide mutation in Duchenne muscular dystrophy, Biosens. Bioelectron. 25 (2010) 2021–2025. [24] K. Iwai, A. Hirono, H. Matsuoka, F. Kawamoto, T. Horie, K. Lin, I.S. Tantular, Y.P. Dachlan, H. Notopuro, N.I. Hidayah, A.M. Salim, H. Fujii, S. Miwa, A. Ishii, Distribution of glucose-6-phosphate dehydrogenase mutations in Southeast Asia, Hum. Genet. 108 (2001) 445–449.
62
Gold nanostructures for mutation detection / N. Seow et al. / Anal. Biochem. 451 (2014) 56–62
[25] M. Hamada, T. Shirakawa, P.S. Lai, K. Nishiyama, S. Uga, M. Matsuo, Two new variants of G6PD deficiencies in Singapore, Nepal Med. Coll. J. 12 (2010) 137–141. [26] J. Turkevich, P.C. Stevenson, J. Hillier, A study of the nucleation and growth processes in the synthesis of colloidal gold, Discuss. Faraday Soc. 11 (1951) 55–75. [27] D.A. Handley, Methods for synthesis of colloidal gold, in: M.A. Hayat (Ed.), Colloidal Gold: Principles, Methods, and Applications, Academic Press, San Diego, 1989, pp. 13–32. [28] S. Link, M.A. El-Sayed, Size and temperature dependence of the plasmon absorption of colloidal gold nanoparticles, J. Phys. Chem. B 103 (1999) 4212– 4217. [29] X. Liu, M. Atwater, J. Wang, Q. Huo, Extinction coefficient of gold nanoparticles with different sizes and different capping ligands, Colloids Surf. B 1 (2007) 3–7. [30] P.K. Jain, K.S. Lee, I.H. El-Sayed, M.A. El-Sayed, Calculated absorption and scattering properties of gold nanoparticles of different size, shape, and composition: applications in biological imaging and biomedicine, J. Phys. Chem. B 110 (2006) 238–248.
[31] X. Ji, X. Song, J. Li, Y. Bai, W. Yang, X. Peng, Size control of gold nanocrystals in citrate reduction: the third role of citrate, J. Am. Chem. Soc. 129 (2007) 13939– 13948. [32] S. Kumar, K.S. Gandhi, R. Kumar, Modeling of formation of gold nanoparticles by citrate method, Ind. Eng. Chem. Res. 46 (2007) 3128–3136. [33] P. Serwer, S.A. Khan, G.A. Griess, Non-denaturing gel–electrophoresis of biological nanoparticles: viruses, J. Chromatogr. A 698 (1995) 251–261. [34] K. Makino, H. Ohshima, Electrophoretic mobility of a colloidal particle with constant surface charge density, Langmuir 26 (2010) 18016–18019. [35] H. Ohshima, Limiting electrophoretic mobility of a highly charged soft particle in an electrolyte solution: solidification effect, J. Colloid Interface Sci. 349 (2010) 641–644. [36] J.M. Nam, C.S. Thaxton, C.A. Mirkin, Nanoparticle-based bio-bar codes for the ultrasensitive detection of proteins, Science 301 (2003) 1884–1886. [37] W. Zhao, M.A. Brook, Y. Li, Design of gold nanoparticle-based colorimetric biosensing assays, ChemBioChem 9 (2008) 2363–2371. [38] W.T. Zhao, I.M. Hsing, Facile and rapid manipulation of DNA surface density on gold nanoparticles using mononucleotide-mediated conjugation, Chem. Commun. 46 (2010) 1314–1316.
Contents lists available at ScienceDirect
Analytical Biochemistry journal homepage: www.elsevier.com/locate/yabio
Gold nanostructures for the multiplex detection of glucose-6-phosphate dehydrogenase gene mutations Nianjia Seow a, Poh San Lai b,⇑, Lin-Yue Lanry Yung a,⇑ a b
Department of Chemical and Biomolecular Engineering, Faculty of Engineering, National University of Singapore, Singapore 119260, Singapore Department of Pediatrics, National University Health System, Yong Loo Lin School of Medicine, National University of Singapore, Singapore 119228, Singapore
a r t i c l e
i n f o
Article history: Received 19 October 2013 Received in revised form 18 January 2014 Accepted 21 January 2014 Available online 1 February 2014 Keywords: Gold nanoparticles Glucose-6-phosphate dehydrogenase (G6PD) deficiency Mutation screening Multiplex detection Gel electrophoresis
a b s t r a c t We describe a gold nanoparticle-based technique for the detection of single-base mutations in the glucose-6-phosphate dehydrogenase (G6PD) gene, a condition that can lead to neonatal jaundice and hemolytic anemia. The aim of this technique is to clearly distinguish different mutations frequently described within the Asian population from their wild-type counterparts and across different mutant variants. Gold nanoparticles of different sizes were synthesized, and each was conjugated with a single-strand DNA (ssDNA) sequence specific for a particular mutation in the G6PD gene. It was found that only mutant targets presented a characteristic band on the agarose gel, indicating the successful formation of dimeric nanostructures. No such dimer bands were observed for the wild-type targets. The difference in the relative dimer band levels allowed different mutant variants to be distinguished from one another. The technique was further validated using G6PD-deficient patient samples. This simple mutation detection method with direct result readout is amenable for rapid and mass screening of samples. Ó 2014 Elsevier Inc. All rights reserved.
Many genetic diseases are caused by single nucleotide point mutations, for example, cystic fibrosis and thalassemia [1]. In particular, glucose-6-phosphate dehydrogenase (G6PD)1 deficiency is a recessive genetic condition arising from single-base mutations in the G6PD gene on the X chromosome and affects 200 to 400 million individuals worldwide [2–4]. It is estimated that 7.5% of global populations carry one or two G6PD mutant alleles, with 2.9% being deficient for this enzyme deficiency. It is also thought that this condition is highly prevalent in malaria endemic countries because it confers resistance against malaria [2]. G6PD-deficient subjects are usually asymptomatic until triggered by infections or the ingestion of certain foods or drugs such as anti-malarial drugs. Clinical outcomes commonly described include neonatal jaundice and acute hemolytic anemia [5]. Thus, there is a need to detect G6PD-deficient subjects in order to avoid triggers for hemolytic anemia and use a substitute medication course when necessary [6]. Routine screening methods typically involve colorimetric biochemical assays. However, ⇑ Corresponding authors. E-mail addresses: [email protected], [email protected] (P.S. Lai), [email protected] (L.-Y.L. Yung). 1 Abbreviations used: G6PD, glucose-6-phosphate dehydrogenase; nAu, gold nanoparticle; PCR, polymerase chain reaction; ssDNA, single-strand DNA; TEM, transmission electron microscopy; HAuCl4, hydrogen tetrachloroaurate(III) trihydrate; Na3Ct, trisodium citrate dehydrate; PPBS, 4,40 -(phenylphosphinidene)bis (benzenesulfonic acid) dipotassium salt hydrate 97%; TBE, Tris–borate–EDTA; UV, ultraviolet; DLS, dynamic light scattering; OEG, oligo(ethylene glycol). http://dx.doi.org/10.1016/j.ab.2014.01.014 0003-2697/Ó 2014 Elsevier Inc. All rights reserved.
although these methods are rapid and relatively cheap, they are qualitative or semi-quantitative and cannot identify the exact mutations [7]. They are also not sensitive enough to detect heterozygotes where both normal and G6PD-deficient erythrocytes are present [8,9]. Molecular diagnosis is possible through identification of the mutations in the G6PD gene. Such mutations have been reported at various locations of the gene in affected individuals, and the dominance of particular mutations is geographically linked. Specific mutation genotypes have been associated with different populations. For example, c.95A>G, c.871G>A, c.1004C>T, c.1024C>T, c.1376G>T, and c.1388G>A mutations are especially common among the Chinese population; these mutations are commonly referred to as the Gaohe, Viangchang, Fushan, Chinese-5, Canton, and Kaiping variants, respectively [10–13]. Typically, molecular diagnosis for any genetic disorder involves either a mutation screening method followed by confirmation through DNA sequencing or directly proceeding to sequencing. This is often laborious due to the experimental steps involved and the sequence data analysis required. A rapid approach for diagnosis is through genotyping for mutations common to a geographical locale, followed by subsequent testing for more uncommon and rare mutations if the former is not informative. Thus, given the clinical relevance and the complement of mutations associated with this condition, there is a motivation to develop a simple and inexpensive multiplex assay with direct result readout for large-scale or population-based screening of G6PD, especially in developing countries where G6PD deficiency is prevalent.
Gold nanostructures for mutation detection / N. Seow et al. / Anal. Biochem. 451 (2014) 56–62
The advent of nanotechnology has led to improvements in molecular techniques in areas such as sensitivity, selectivity, and immediacy of readouts. Nanoparticles offer unique properties, such as surface plasmon resonance, and are amenable to biofunctionalizations, making them ideal as labeling tags or readout platforms [14–16]. Mirkin and coworkers have developed various colorimetric assays that leverage on the spectral shift of gold nanoparticles (nAu) as they aggregate in the presence of DNA targets [17]. The readouts can easily be visualized by the naked eye. Further application using bio-barcode assay even offers polymerase chain reaction (PCR)-like performance with a zeptomolar (zM) detection limit [17,18]. Control in the functionalization of nAu was also described by Alivisatos and coworkers as they showed the formation of discrete nAu structures bearing a distinct number of single-strand DNA (ssDNA) that is evidenced in transmission electron microscopy (TEM) and gel platforms [19,20]. This allows the formation of distinct nanostructures, as Qin and Yung demonstrated through the use of nAu dimers to discriminate single-nucleotide mutations in diseases such as Duchenne muscular dystrophy [21–23]. Here we report the use of different nAu probes for detecting four characteristic G6PD mutations prevalent in the Asian population, namely Canton, Mahidol, Union, and A+ variants [3,24,25]. These probes are not only specific for a perfectly matched mutant target relative to the wild-type nontarget, but when used in tandem they are also able to discriminate variants in the G6PD gene. This multiplex readout is presented in a direct and unambiguous manner, which gives a yes/no readout for the variant nucleotide of interest and also discriminates among different single-base mutations through the differential mobility of the probes in agarose gel electrophoresis. Clinical samples were probed with different types and sizes of probes, and results had conformed to predicted expectations. The success of this technique has wider implications in the screening for disease-associated single-nucleotide mutations and polymorphisms and for biomarker discovery in general. Materials and methods Materials Hydrogen tetrachloroaurate(III) trihydrate (HAuCl4), trisodium citrate dihydrate (Na3Ct), tannic acid, 4,40 -(phenylphosphinidene)bis (benzenesulfonic acid) dipotassium salt hydrate 97% (PPBS), NaCl, MgCl2, and Tris–borate–EDTA (TBE) buffer were purchased from Sigma–Aldrich (Singapore). Tris buffer (pH 8.0) and agarose powder were purchased from 1st Base (Singapore). Modified synthetic DNA molecules (modified with or without 30 or 50 thiol end group) were purchased from 1st Base Custom Oligos (Singapore). Milli-Q water with resistance > 18 MO/cm was used throughout the experiments. For the amplification of patient samples, standard PCR conditions were applied using a thermocycler (Biometra, Germany): initial denaturation at 95 °C for 7 min; followed by 35 cycles of 95 °C for 1 min, 58 °C for 1 min, and 72 °C for 3 min; and a final extension at 72 °C for 7 min. Amplifications were performed in total reaction volumes of 25 ll containing 100 ng of DNA, 0.4 lM forward and reverse primers, 1 U of Taq DNA Polymerase, 0.4 mM deoxynucleotide triphosphates (dNTPs), 2 mM MgCl2, and 1 buffer (Promega, Singapore). All samples were obtained following approved institutional ethics guidelines, and the study was approved by the institutional review board (IRB) of the National University of Singapore.
57
precursor, synthesis time, and temperature) were kept constant, whereas concentration of reductant (either Na3Ct only or both Na3Ct and tannic acid) was varied [27,28]. nAu products were characterized first with ultraviolet (UV)–visible spectroscopy and TEM (JEM-2100F, 200 kV, Joel). The TEM micrographs were analyzed with ImageJ software (National Institutes of Health) for size measurement (P200 particles). The particle size distribution was further characterized by determining particles’ electrophoretic mobilities on a 3% agarose gel (5 V/cm) and dynamic light scattering (DLS) (Zetasizer Nano ZS, Malvern). Fabrication of nAu–ssDNA conjugate probes Surface passivation of nAu with PPBS was first done for at least 4 h, followed by incubation with ssDNA specific for the mutant targets. Here, 100- and 18-base ssDNA–nAu conjugate probes for each target were prepared (the gene sequences of all the ssDNA can be found in the online Supplementary material). Next, 3 h after the oligo addition, thiolated oligo(ethylene glycol) (OEG) was added at a 500:1 OEG/nAu ratio. The nAu–ssDNA probes were prepared in 75 to 200 mM NaCl initially (depending on the size of nAu), and gradually the NaCl concentration was raised to between 350 and 600 mM (larger nAu conjugates are generally less stable at higher NaCl conditions). The probes were subsequently washed in 50 mM Tris and NaCl. The 100-base probes were run on a 3% agarose gel, and conjugate monomers (probes bearing a single ssDNA strand) were recovered. The 18-base probes were used without prior purification. The concentrations of probes were estimated from their UV absorbance [29,30]. Prior to use, all probes were screened using UV spectroscopy and DLS to ensure that they remained monodispersed and unaggregated. Probes without targets were also used as controls in subsequent tests and gel runs, and their monodispersity was further confirmed by the (single) band observed. Formation of dimeric nanostructures in the presence of mutant targets Based on our previous work, and after optimization for different sizes of nAu, a probe/target ratio ranging from 1:1 to 1:1.5 was used. By default, 1 pmol of synthetic DNA was used as a target for all of the runs. Hybridization conditions were kept at 50 mM Tris buffer (pH 8.0), 100 mM NaCl, and 2 mM MgCl2, with the samples being heated up to 75 °C and cooled by 0.5 °C every 30 s. The hybridized samples were left to stand for at least 1 h. The samples were then characterized using a 3% agarose gel (75 V for 45 min at 4 °C). Gel images were taken using the Syngene GeneGenius UV/ white light gel documentation system. The gel images, with a 10% enhancement of contrast, were subsequently analyzed using ImageJ software, and the density of the dimer bands was measured relative to that of the unbound probes control. This densitometric analysis provided more quantitative information on the dimer bands and further confirmed the presence of the target. A schematic of the nanostructure formation and gel readout is shown in Fig. 1, where two nAu–ssDNA probes (carrying 100and 18-base ssDNA, respectively) hybridized to a mutant target sequence and formed a dimeric gold nanostructure. The successful formation of the structure was visualized by agarose gel electrophoresis, where a distinct band could be readily seen by the naked eye. Any sequence mismatches would not lead to the formation of such dimer bands. Results
Synthesis and characterization of gold nanoparticles
Characterization of different sizes of gold nanoparticles
The synthesis of nAu was based on the Turkevich protocol [26] for the synthesis of colloidal gold. Reaction conditions (of HAuCl4
The combination of water, Na3Ct, and HAuCl4 is integral to the successful synthesis of nAu of the desired size(s) because varying
58
Gold nanostructures for mutation detection / N. Seow et al. / Anal. Biochem. 451 (2014) 56–62
wells
Dimers
& Unbound probes
Probes only
Probes with perfectmatched target
Fig.1. Design of nAu–ssDNA probe system. A 100- and 18-base probe tandem would bind onto the same single-stranded target to form a dimer. The mutation site is located on the 18-base probe. The 3% agarose gel image shows the formation of dimers in the presence of a perfect matched target. In the absence of targets (probes only), no such dimer bands were observed.
the Na3Ct/HAuCl4 ratio changes the reaction pathway and, in turn, affects the quality of the nAu product. Peng and coworkers reported that, besides acting as a reducing and capping agent, citrate is a pH mediator, and controlling its addition into the reaction mixture can have a direct effect on the synthesis outcome [31]. In addition, fixing the Na3Ct/HAuCl4 ratio at approximately 3 produced the smallest nanoparticle size. When the Na3Ct/HAuCl4 ratio decreased from 3 or below, the product size was progressively larger because the conditions promoted the growth of nanocrystals in the traditional route of nucleation followed by diffusion-controlled net growth [32]. When the ratio increased above 3, there was a corresponding increase in product size, although the increase was more gradual. This could be attributed to the high citrate concentration, which caused the reaction pH to be lower than the pH 6.5 switch point. As a result, there were changes to the structure and reactivity of the Au(III) complexes, with the reaction pathway instead involving nucleation, aggregation, and intraparticle ripening, which led to larger sized nAu. Given that we were leveraging on the properties of different nAu sizes, it was imperative that we could properly control the sizes of nAu synthesized and replicate them with ease. Hence, we fixed the starting concentration of HAuCl4 at 0.1 mg/ml and varied the concentration of Na3Ct at a fixed total reaction volume and synthesis conditions such as temperature and incubation time. Na3Ct was at least five times in excess of HAuCl4. Through controlling the relative concentrations of HAuCl4 precursor and reductant, the resulting nAu showed a size range of 15 to 30 nm. To achieve nAu sizes of 12 nm or lower, a combination of Na3Ct and tannic acid was used, with 9-nm nAu being the smallest synthesized. Size measurement via TEM and DLS further verified the results (Fig. 2A and B). The synthesized nAu showed a narrow size distribution, with the standard deviation within 10% of the average particle size. The DLS readout also showed that each synthesized size presented a peak position distinct from that of the other sizes. Under agarose gel electrophoresis, the as-synthesized nAu of different sizes exhibited the mobility corresponding to particle size, with larger particles showing slower mobilities than their smaller sized counterparts (Fig. 2A). The mobility differentiation among all particle sizes subsequently allowed us to use electrophoresis as the readout platform. The size range (9–25 nm) was also adequate for a multiplex detection scheme because the mobilities of different size bands were visually distinguishable. Formation of discrete nanostructures with different sizes of gold nanoparticles We first used the Canton G6PD point mutation, which showed a G > T mutation, as a proof of principle for the methodology. The aim was to investigate the ability of the nAu probes to form dimers
as well as the relative mobilities of the nanostructures. Different sizes of nAu (8.5, 11, 14, 20, and 25 nm) were conjugated to ssDNA specific to the mutant Canton variant. Probes bearing a single DNA strand were recovered from the agarose gel to ensure dimer formation and no other nanostructures. After optimization of hybridization conditions for each of the sizes, dimers were successfully formed and presented as distinct dimeric bands on the gel platform (Fig. 3). The 8.5-nm 18-base and 25-nm 100-base probes were used as the upper and lower limit controls because the former would show the greatest mobility and the latter would show the least mobility among the probes. Thus, they served as reference points for evaluating the mobilities of the unbound probes and dimer bands when targets were present. All of the unbound probe bands were localized within these two reference bands, with the dimer bands showing a slower mobility behind the unbound probe bands. Potentially, the formation of dimers not only would provide a yes/no readout with regard to the presence of a DNA target but also would discriminate between different mutations due to the respective dimers being presented at distinct locations on the gel. Higher order bands that migrate at slower mobility than the dimer bands were formed due to multiple ssDNA strands conjugated on the 18-base nAu probes. Because 18-base ssDNA is too short to induce sufficient electrophoretic mobility, nAu probes with different numbers of ssDNA strands could not be resolved into a distinct gel band during nAu–ssDNA probe fabrication; thus, no purification was done prior to use. Instead, the number of 18-base ssDNA on the 18-base probes was adjusted only through careful stoichiometric control, such that dimers would still be the dominant population in the presence of complementary target DNA, but the presence of dimer or higher order -mer bands was still indicative of the presence of target DNA. Multiplex detection with discrimination between mutant and wildtype targets and across different point mutations nAu–ssDNA probes for the four different variants (Union, Mahidol, Canton, and A+) were fabricated, and the dimeric nanostructures reflecting positive readouts were found to exhibit distinct electrophoretic mobilities. As seen in Fig. 4, the four variants were queried over eight nAu sizes, and dimer bands were observed only for the mutant targets to which the probes were perfectly complementary. The wild-type sequences that contained a single base difference did not produce any dimer bands. The mismatch site, being located on the shorter 18-base nAu probe, was affected by the electrostatic–steric effect exerted by the nearby nAu, and this destabilized the complementary binding of the probe for the target, which had an overlap of only 13 bases. Despite different sizes of nAu exhibiting varied surface charge densities, the electrostatic–steric effect was still significant in bringing about the desired discrimination property. A single-base mismatch was enough to destabilize and dissemble any nanostructure formed. More importantly, all of the dimer bands were observed at different positions on the gel that were distinct from one another. Although the probes carried ssDNA of the same length, the difference in the nAu sizes for different mutant targets resulted in the differential mobilities observed. This was attributed to the mass being an important property that controlled the particle mobility on a gel platform, with the larger nAu probes exhibiting slower dimer movement [33–35]. This provided an added level of discrimination for different sequences because each was expected to have a unique position based on the sizes of the nAu particles. In this particular case, the spatial multiplex detection of different G6PD variants discriminated a mutant sequence not only from its wild-type counterpart but also across different point mutations through a simple and direct observation of the mobilities of the nAu probe across lanes. A positive readout would be expected in relation to
Gold nanostructures for mutation detection / N. Seow et al. / Anal. Biochem. 451 (2014) 56–62
59
Fig.2. (A) TEM micrographs and gel image of six different sizes of nAu—9.0 nm (a), 12.3 nm (b), 15.2 nm (c), 17.3 nm (d), 21.0 nm (e), and 27.2 nm (f)—synthesized with different amounts of citrate and tannic acid reductants. (B) Particle distribution of different sizes of nAu using DLS. All of the particles show distinct size distributions and electrophoretic mobilities.
8.5nm dimer
11nm dimer
14nm dimer
20nm dimer 25nm dimer
25nm 100-base probe (lower limit control) 8.5nm 18-base probe (upper limit control)
denotes the dimermonomer separation
Direction of migration
Fig.3. Dimer formation with 8.5-, 11-, 14-, 20-, and 25-nm nAu (with Canton variant as the model) in 3% agarose gel (75 V, 60 min). Target loading was 1 pmol, and the amounts of Canton mutation-specific probes used were varied to give optimal observation of the dimer band. Higher order multimer bands were attributed to the unpurified 18-base probes, some of which may carry more than one ssDNA per nAu and led to the formation of not just dimer but also trimer, tetramer, and so forth.
60
Gold nanostructures for mutation detection / N. Seow et al. / Anal. Biochem. 451 (2014) 56–62
Fig.4. Multiplex detection of four variants (Union, Mahidol, Canton, and A+) complement on a 3% agarose gel. For each mutation type, the same quantity of the probes specific to the particular mutation was used to detect mutant and wild-type targets that showed a single-base difference. The appearance of dimer bands on the gel (or the lack thereof) provides identification to the type of target present. The dimer band for each variant shows distinct mobility from the others due to the different sizes of nAu used, providing an advantage in discriminating target detection during multiplex screening of varied G6PD mutations. mut: mutant (perfectly matched) target, expected to produce dimer bands; wt: wild-type (mismatched) target, not expected to produce dimer bands.
the presence of the specific dimer bands as well as its relative position to its counterparts. Querying of clinical samples using different sized probes To validate the above approach, clinical samples from G6PDdeficient patients were analyzed using the different types of nAu probes. Four different patient samples—two with the Mahidol mutation (M1 and M2) and two with Canton mutation (C3 and C4)—were obtained with informed consent and IRB approval. The genomic DNA samples were PCR amplified, purified, and subsequently detected by the respective Mahidol and Canton probes. From Fig. 5, when queried with their specific probes, the Mahidol (with 11-nm probes) and Canton (with 15-, 17-, and 20-nm probes) patient samples all exhibited dimer bands. When the target was absent, no dimer band was observed. The observations were also semi-quantified via the density of the bands, which took into account the intensity and area of the bands, and the dimer bands’ densities were approximately 10 to 20% that of the unbound probes. The densitometric analysis provided a semi-quantitative assay for confirming the presence of the dimers (and target). The dimer bands, although less dense than their synthetic target counterparts, were easily detected by the naked eye without the need for additional instruments. Fainter bands were observed because the clinical PCR-amplified samples were long duplex DNA, unlike the single-stranded synthetic DNA targets used in the previous sections. Given that the duplexes are typically more than 120 bases in
11nm Mahidol probes Synthetic Mahidol Patient M1 Patient M2 (Mahidol (Mahidol target mutation) mutation) control
15nm Canton probes Patient C3 (Canton mutation)
Patient C4 (Canton mutation)
length, whereas the 18-base probes have only 13 complementary bases to the target, the tendency for the target duplex to rehybridize with its complementary strand is higher. Thus, it was necessary to include an annealing step, such that duplexes would dehybridize and the target strands would be available for binding with the nAu probes, and all four clinical samples were successfully detected. We have shown that different sizes of probes, ranging from 11 to 20 nm, were equally capable of detecting the Canton and Mahidol clinical samples, hence validating the multiplex detection proposed. Potentially, this can be a useful technique for easy and direct screening, especially at the clinical setting for precise detection of common G6PD mutations. Discussion The development of nAu-based diagnostic techniques hinges much on the understanding of the nature and properties of nAu and their surface chemistries and, finally, fabricating the nAubased probes for the desired applications. For example, the desired surface loading of DNA onto the nanoparticle differs depending on the type of application, whereas care is taken to ensure the stability of nanoparticle dispersion and kinetics of DNA binding onto the nAu surface. Typically, high DNA-loaded conjugates are used for amplification purposes and aggregation assays [36]. On the other hand, ssDNA–nAu conjugates with low ssDNA loading have shown much promise as building blocks in complex nanostructure formation. However, the need for a controlled loading of ssDNA poses a
17nm Canton probes Patient C3
Patient C4
20nm Canton probes Patient C3
Patient C4
Control (no target)
Dimers (probes bound to target)
Unbound probes Dimers expected (yes/no) Dimers formed (yes/no) Relative intensity to unbound probes
yes
yes
yes
yes
yes
yes
yes
yes
no
yes
yes
yes
yes
yes
yes
yes
yes
no
0.17
0.13
0.18
0.09
0.09
0.21
0.12
0.15
0
Direction of migration
Fig.5. PCR-amplified samples from four patients—two with Mahidol mutation (M1 and M2) and two with Canton mutation (C3 and C4)—were received. The patient samples were queried with the respective mutation-specific probes (11-nm Mahidol probes for M1 and M2; 15-, 17-, and 20-nm Canton probes for C3 and C4). All of the patient samples exhibited dimer bands when tested with their respective probes, which followed predicted expectations. The readout is further characterized via the density of the dimer bands relative to the unbound probes. In the absence of targets, no dimers were observed.
Gold nanostructures for mutation detection / N. Seow et al. / Anal. Biochem. 451 (2014) 56–62
challenge to probe fabrication, which affects the formation of distinct nanostructures and their use in plasmonic and electronic applications. Gel electrophoresis and anion-exchange high-performance liquid chromatography (HPLC) have been the foremost choices to separate conjugates bearing different numbers of ssDNA due to their compatible resolution and ease of use. In particular, the gel platform was found to be suitable not just for the purification and subsequent recovery of conjugates bearing distinct numbers of ssDNA but also for the visualization of different nanostructures formed achieved in this work. Much care was taken during the probe fabrication process because inherently nAu are sensitive to environmental conditions such as Na+ levels, which might bring forth aggregation. Thus, for the timely fabrication of the desired conjugate probes, addition of NaCl was carefully done because it is paradoxically known that a higher NaCl level can enhance DNA binding onto the nAu surface while at the same time promoting clustering of nAu and eventual aggregation, handicapping the nanostructure formation [37]. In our case, the ssDNA conjugation protocol was optimized for all of the nAu sizes used, especially the larger sized nAu particles that were more susceptible to aggregation at high salt loadings relative to their smaller sized counterparts. This ensured the efficient hybridization of probes to the targets without interference of salt-induced clustering of probes or formation of nonspecific nanostructures. Contrary to our previous work, where thiolated 5-thymidine (5T) was used in the preparation of the conjugate probes, this study showed that OEG was as good a passivating agent as 5T while being only a tenth in cost. OEG was able to compete with thiolated ssDNA and allowed both the 100- and 18-base ssDNA coverage on the nAu to be kept low, thereby favoring the formation of discrete conjugates with low ssDNA loading [38]. From the hybridization studies, probe–target binding was just as effective, which suggested that the OEG-coated surface was non-interacting and kept the probe ssDNA extended outward and not collapsed onto the nAu surface, hence facilitating the hybridization process. Although we tested eight nAu sizes in the 7- to 25-nm range, it is possible to include additional nAu sizes as small as 5 nm and as large as 40 nm, with each size being used to detect for a specific mutant variant of the G6PD gene. This will be especially useful to simultaneously cover screening of other less common point mutations in the G6PD gene. This, coupled with the developed procedure of probe fabrication, target hybridization, and visualization as described in this study, offers a potentially powerful multiplex detection method for G6PD screening of known mutations in a population. As compared with conventional molecular techniques and other nanoparticle-based platforms, gold nanostructure probes provide a simple and direct manner of detection that can be applied to the detection of point mutations in G6PD deficiency as well as for other disorders.
Conclusion Given the need for a simple and direct system for the detection of multiple nucleic acid targets, we have developed an nAu system that, through the formation of distinct structures from different sizes of nAu, can be clearly visualized on a gel platform. Wild-type and single-base mutant sequences were discriminated, as well as across variants, and this technique was further validated using clinical samples. The ability to detect the readouts from the nAu hybridization by direct visualization from a gel-based assay makes this a simple and inexpensive method. Our method can potentially be applied for screening other pathogenic point mutations in disease genes as well as for genotyping common single-nucleotide polymorphisms used as markers for disease association studies in disease gene discovery.
61
Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.ab.2014.01.014.
References [1] M.N. Kuppuswamy, J.W. Hoffmann, C.K. Kasper, S.G. Spitze, S.L. Groce, S.P. Bajaj, Single nucleotide primer extension to detect genetic diseases: experimental application to hemophilia B (factor IX) and cystic fibrosis genes, Proc. Natl. Acad. Sci. USA 88 (1991) 1143–1147. [2] M.D. Cappellini, G. Fiorelli, Glucose-6-phosphate dehydrogenase deficiency, Lancet 371 (2008) 64–74. [3] P. Delbini, L. Duca, I. Nava, D. Tavazzi, M.D. Cappellini, Gene symbol: G6PD. Disease: glucose-6-phosphate dehydrogenase deficiency, Hum. Genet. 124 (2008) 312–313. [4] P.S. Lai, T. Shirakawa, K. Nishiyama, M. Matsuo, R. Joseph, S.H. Quak, G6PD deficiency and application of the MPTP technique, in: P.S. Lai, E.P.H. Yap (Eds.), Frontiers of Human Genetics: Diseases and Technologies, World Scientific, Singapore, 2001, pp. 121–138. [5] J.E. Frank, Diagnosis and management of G6PD deficiency, Am. Fam. Physician 72 (2005) 1277–1282. [6] W.E. Evans, M.V. Relling, Pharmacogenomics: translating functional genomics into rational therapeutics, Science 286 (1999) 487–491. [7] I.S. Tantular, F. Kawamoto, An improved, simple screening method for detection of glucose-6-phosphate dehydrogenase deficiency, Trop. Med. Int. Health 8 (2003) 569–574. [8] G.J. Reclos, C.J. Hatzidakis, K.H. Schulpis, Glucose-6-phosphate dehydrogenase deficiency neonatal screening: preliminary evidence that a high percentage of partially deficient female neonates are missed during routine screening, J. Med. Screen. 7 (2000) 46–51. [9] O. Ainoon, A. Alawiyah, Y.H. Yu, Semiquantitative screening test for G6PD deficiency detects severe deficiency but misses a substantial proportion of partially-deficient females, Southeast Asian, J. Trop. Med. Public Health 34 (2003) 405–414. [10] L. Li, Y.Q. Zou, Q.Z. Xiao, T.Z. Yan, X.M. Xu, Development and evaluation of a reverse dot blot assay for the simultaneous detection of six common Chinese G6PD mutations and one polymorphism, Blood Cells Mol. Dis. 41 (2008) 17– 21. [11] A. Minucci, L. Gentile, C. Zuppi, B. Giardina, E. Capoluongo, Rapid and simple identification of the commonest glucose-6-phosphate dehydrogenase (G6PD) Italian mutations: from DNA extraction to genotyping, Clin. Chim. Acta 413 (2012) 1018–1019. [12] A. Minucci, P. Concolino, M. Antenucci, C. Santonocito, F. Ameglio, C. Zuppi, B. Giardina, E. Capoluongo, Description of a novel missense mutation of glucose6-phosphate dehydrogenase gene associated with asymptomatic high enzyme deficiency, Clin. Biochem. 40 (2007) 856–858. [13] T. Yan, R. Cai, O. Mo, D. Zhu, H. Ouyang, F. Huang, L. Li, X. Liang, X. Xu, Incidence and complete molecular characterization of glucose-6-phosphate dehydrogenase deficiency in the Guangxi Zhuang autonomous region of southern China: description of four novel mutations, Haematologica 91 (2006) 1321–1328. [14] S.S. Agasti, S. Rana, M.H. Park, C.K. Kim, C.C. You, V.M. Rotello, Nanoparticles for detection and diagnosis, Adv. Drug Deliv. Rev. 62 (2010) 316–328. [15] J.N. Anker, W.P. Hall, O. Lyandres, N.C. Shah, J. Zhao, R.P. Van Duyne, Biosensing with plasmonic nanosensors, Nat. Mater. 7 (2008) 442–453. [16] Z. Zhan, X. Ma, C. Cao, S.J. Sim, Gold-based optical biosensor for singlemismatched DNA detection using salt-induced hybridization, Biosens. Bioelectron. 32 (2012) 127–132. [17] M.S. Han, A.K.R. Lytton-Jean, B.K. Oh, J. Heo, C.A. Mirkin, Colorimetric screening of DNA-binding molecules with gold nanoparticle probes, Angew. Chem. Int. Ed. Engl. 45 (2006) 1807–1810. [18] J.M. Nam, S.I. Stoeva, C.A. Mirkin, Bio-bar-code-based DNA detection with PCRlike sensitivity, J. Am. Chem. Soc. 126 (2004) 5932–5933. [19] D. Zanchet, C.M. Micheel, W.J. Parak, D. Gerion, S.C. Williams, A.P. Alivisatos, Electrophoretic and structural studies of DNA-directed Au nanoparticle groupings, J. Phys. Chem. B 106 (2002) 11758–11763. [20] D. Zanchet, C.M. Micheel, W.J. Parak, D. Gerion, A.J. Alivisatos, Electrophoretic isolation of discrete Au nanocrystal/DNA conjugates, Nano Lett. 1 (2001) 32– 35. [21] W.J. Qin, L.Y.L. Yung, Nanoparticle–DNA conjugates bearing a specific number of short DNA strands by enzymatic manipulation of nanoparticle-bound DNA, Langmuir 21 (2005) 11330–11334. [22] W.J. Qin, L.Y.L. Yung, Nanoparticle-based detection and quantification of DNA with single nucleotide polymorphism (SNP) discrimination selectivity, Nucleic Acids Res. 35 (2007) e111. [23] W.J. Qin, O.S. Yim, P.S. Lai, L.Y.L. Yung, Dimeric gold assembly for detection and discrimination of single nucleotide mutation in Duchenne muscular dystrophy, Biosens. Bioelectron. 25 (2010) 2021–2025. [24] K. Iwai, A. Hirono, H. Matsuoka, F. Kawamoto, T. Horie, K. Lin, I.S. Tantular, Y.P. Dachlan, H. Notopuro, N.I. Hidayah, A.M. Salim, H. Fujii, S. Miwa, A. Ishii, Distribution of glucose-6-phosphate dehydrogenase mutations in Southeast Asia, Hum. Genet. 108 (2001) 445–449.
62
Gold nanostructures for mutation detection / N. Seow et al. / Anal. Biochem. 451 (2014) 56–62
[25] M. Hamada, T. Shirakawa, P.S. Lai, K. Nishiyama, S. Uga, M. Matsuo, Two new variants of G6PD deficiencies in Singapore, Nepal Med. Coll. J. 12 (2010) 137–141. [26] J. Turkevich, P.C. Stevenson, J. Hillier, A study of the nucleation and growth processes in the synthesis of colloidal gold, Discuss. Faraday Soc. 11 (1951) 55–75. [27] D.A. Handley, Methods for synthesis of colloidal gold, in: M.A. Hayat (Ed.), Colloidal Gold: Principles, Methods, and Applications, Academic Press, San Diego, 1989, pp. 13–32. [28] S. Link, M.A. El-Sayed, Size and temperature dependence of the plasmon absorption of colloidal gold nanoparticles, J. Phys. Chem. B 103 (1999) 4212– 4217. [29] X. Liu, M. Atwater, J. Wang, Q. Huo, Extinction coefficient of gold nanoparticles with different sizes and different capping ligands, Colloids Surf. B 1 (2007) 3–7. [30] P.K. Jain, K.S. Lee, I.H. El-Sayed, M.A. El-Sayed, Calculated absorption and scattering properties of gold nanoparticles of different size, shape, and composition: applications in biological imaging and biomedicine, J. Phys. Chem. B 110 (2006) 238–248.
[31] X. Ji, X. Song, J. Li, Y. Bai, W. Yang, X. Peng, Size control of gold nanocrystals in citrate reduction: the third role of citrate, J. Am. Chem. Soc. 129 (2007) 13939– 13948. [32] S. Kumar, K.S. Gandhi, R. Kumar, Modeling of formation of gold nanoparticles by citrate method, Ind. Eng. Chem. Res. 46 (2007) 3128–3136. [33] P. Serwer, S.A. Khan, G.A. Griess, Non-denaturing gel–electrophoresis of biological nanoparticles: viruses, J. Chromatogr. A 698 (1995) 251–261. [34] K. Makino, H. Ohshima, Electrophoretic mobility of a colloidal particle with constant surface charge density, Langmuir 26 (2010) 18016–18019. [35] H. Ohshima, Limiting electrophoretic mobility of a highly charged soft particle in an electrolyte solution: solidification effect, J. Colloid Interface Sci. 349 (2010) 641–644. [36] J.M. Nam, C.S. Thaxton, C.A. Mirkin, Nanoparticle-based bio-bar codes for the ultrasensitive detection of proteins, Science 301 (2003) 1884–1886. [37] W. Zhao, M.A. Brook, Y. Li, Design of gold nanoparticle-based colorimetric biosensing assays, ChemBioChem 9 (2008) 2363–2371. [38] W.T. Zhao, I.M. Hsing, Facile and rapid manipulation of DNA surface density on gold nanoparticles using mononucleotide-mediated conjugation, Chem. Commun. 46 (2010) 1314–1316.
